Predicting Patent Litigation

Abstract

Patent lawsuits are disruptive, unpredictable, and costly. The inability to anticipate patent litigation makes it practically uninsurable, exposes companies to late-stage suits, and drives companies to rapidly accumulate patents in order to ward off litigation. This article confronts this systemic problem, by examining the factors that lead a particular patent to be litigated – only around 1% of patents ever is. It relates the eventual litigation of a patent to earlier events in the patent’s life, including changes in ownership of the patent (assignments, transfers, and changes in owner size), continued investment in the patent (reexamination, maintenance fees), securitization of the patent, and citations to the patent. To date these “acquired” characteristics, developed after a patent has issued in contrast to the intrinsic qualities a patent is “born” with, have been the subject of limited academic study.

The results are dramatic: along all the dimensions studied, patents that end up in litigation have markedly different characteristics than patents that don’t. Importantly for predictive purposes, these differences develop prior to the time of litigation, suggesting that the riskiest patents can be identified ahead of time. The results are also surprising, showing that the likelihood of litigation depends on not only how valuable the patent is but also its owner and transaction history. They draw attention to a policy area that has been long overlooked – ensuring that the public has notice not only of what a patent covers, but also of who owns it and what happens to it. The ease with which patent owners can hide who they are and what they are doing with their patents raise cause for concern, and potential reform, of the patent system.

Keywords: patent, empirical, patent litigation, patent notice, patent transactions